feat: Unit tests for serialize_v0 file model

.github/workflows/unit_tests.yml

@@ -0,0 +1,32 @@
+name: Run unit tests
+on:
+  pull_request:
+    branches: [main]
+    types: [opened, synchronize]
+
+permissions: {}
+
+jobs:
+  model-signing-unit-tests:
+    name: Run unit tests for signing
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1
+    - uses: actions/setup-python@65d7f2d534ac1bc67fcd62888c5f4f3d2cb2b236 # v4.7.1
+      with:
+        python-version: 3.11
+        cache: pip
+        cache-dependency-path: |
+          model_signing/install/requirements.txt
+          model_signing/install/requirements.test.txt
+    - name: Install dependencies
+      run: |
+        set -euo pipefail
+        python -m venv venv
+        source venv/bin/activate
+        python -m pip install --require-hashes -r model_signing/install/requirements.txt
+        python -m pip install --require-hashes -r model_signing/install/requirements.test.txt
+    - name: Run unit tests
+      run: |
+        set -euo pipefail
+        pytest .

model_signing/install/requirements.test.txt

@@ -0,0 +1,22 @@
+#
+# This file is autogenerated by pip-compile with Python 3.11
+# by the following command:
+#
+#    pip-compile --allow-unsafe --generate-hashes --output-file=install/requirements.test.txt install/requirements.test.in
+#
+iniconfig==2.0.0 \
+    --hash=sha256:2d91e135bf72d31a410b17c16da610a82cb55f6b0477d1a902134b24a455b8b3 \
+    --hash=sha256:b6a85871a79d2e3b22d2d1b94ac2824226a63c6b741c88f7ae975f18b6778374
+    # via pytest
+packaging==23.2 \
+    --hash=sha256:048fb0e9405036518eaaf48a55953c750c11e1a1b68e0dd1a9d62ed0c092cfc5 \
+    --hash=sha256:8c491190033a9af7e1d931d0b5dacc2ef47509b34dd0de67ed209b5203fc88c7
+    # via pytest
+pluggy==1.3.0 \
+    --hash=sha256:cf61ae8f126ac6f7c451172cf30e3e43d3ca77615509771b3a984a0730651e12 \
+    --hash=sha256:d89c696a773f8bd377d18e5ecda92b7a3793cbe66c87060a6fb58c7b6e1061f7
+    # via pytest
+pytest==7.4.3 \
+    --hash=sha256:0d009c083ea859a71b76adf7c1d502e4bc170b80a8ef002da5806527b9591fac \
+    --hash=sha256:d989d136982de4e3b29dabcc838ad581c64e8ed52c11fbe86ddebd9da0818cd5
+    # via -r install/requirements.test.in

model_signing/serialize.py

@@ -281,7 +281,7 @@ def task(task_info: []):
     def serialize_v1(path: Path, chunk: int, signature_path: Path,
                      ignorepaths: [Path] = []) -> bytes:
         if not path.exists():
-            raise ValueError(f"{str(path)} does not eixst")
+            raise ValueError(f"{str(path)} does not exist")
 
         if not allow_symlinks and path.is_symlink():
             raise ValueError(f"{str(path)} is a symlink")
@@ -312,7 +312,7 @@ def serialize_v1(path: Path, chunk: int, signature_path: Path,
     def serialize_v0(path: Path, chunk: int, signature_path: Path,
                      ignorepaths: [Path] = []) -> bytes:
         if not path.exists():
-            raise ValueError(f"{str(path)} does not eixst")
+            raise ValueError(f"{str(path)} does not exist")
 
         if not allow_symlinks and path.is_symlink():
             raise ValueError(f"{str(path)} is a symlink")

model_signing/serialize_test.py

@@ -0,0 +1,140 @@
+import os
+from pathlib import Path
+import pytest
+from serialize import Serializer
+import shutil
+
+
+testdata_dir = "testdata"
+
+
+# Utility functions.
+def create_folder(name: str) -> Path:
+    p = os.path.join(os.getcwd(), testdata_dir, name)
+    os.makedirs(p)
+    return Path(p)
+
+
+def create_symlinks(src: str, dst: str) -> Path:
+    psrc = os.path.join(os.getcwd(), testdata_dir, src)
+    pdst = os.path.join(os.getcwd(), testdata_dir, dst)
+    os.symlink(psrc, pdst)
+    return Path(dst)
+
+
+def cleanup_model(p: Path) -> None:
+    if p.is_dir():
+        shutil.rmtree(p)
+    elif p.is_file():
+        os.unlink(p)
+        try:
+            os.unlink(p.with_suffix(".sig"))
+        except FileNotFoundError:
+            pass
+
+
+def create_file(name: str, data: bytes) -> Path:
+    p = os.path.join(os.getcwd(), testdata_dir, name)
+    with open(p, "wb") as f:
+        f.write(data)
+    return Path(p)
+
+
+def create_random_file(name: str, size: int) -> (Path, bytes):
+    p = os.path.join(os.getcwd(), testdata_dir, name)
+    content = os.urandom(size)
+    with open(p, "wb") as f:
+        f.write(content)
+    return Path(p), content
+
+
+def signature_path(model: Path) -> Path:
+    if model.is_file():
+        return model.with_suffix(".sig")
+    return model.joinpath("model.sig")
+
+
+class Test_serialize_v0:
+    # symlink in root folder raises ValueError exception.
+    def test_symlink_root(self):
+        folder = "with_root_symlinks"
+        model = create_folder(folder)
+        sig = signature_path(model)
+        create_symlinks(".", os.path.join(folder, "root_link"))
+        with pytest.raises(ValueError):
+            Serializer.serialize_v0(Path(folder), 0, sig)
+        cleanup_model(model)
+
+    # symlink in non-root folder raises ValueError exception.
+    def test_symlink_nonroot(self):
+        model = create_folder("with_nonroot_symlinks")
+        sub_folder = model.joinpath("sub")
+        create_folder(str(sub_folder))
+        sig = signature_path(model)
+        create_symlinks(".", os.path.join(sub_folder, "sub_link"))
+        with pytest.raises(ValueError):
+            Serializer.serialize_v0(model, 0, sig)
+        cleanup_model(model)
+
+    # File serialization works.
+    def test_known_file(self):
+        file = "model_file"
+        data = b"hellow world content"
+        model = create_file(file, data)
+        sig_path = signature_path(model)
+        expected = b'x\x9d\xa4N\x9f\xeajd\xd8\x87\x84\x1a\xd3\xb3\xfc\xeb\xf6\r\x01\x9fi8#\xd8qU\x90\xca\x9d\x83\xe1\x8b'  # noqa: E501 ignore long line warning
+        computed = Serializer.serialize_v0(model, 0, sig_path)
+        assert (computed == expected)
+        cleanup_model(model)
+
+    # File serialization returns the same results for different chunk sizes.
+    def test_file_chuncks(self):
+        file = "model_file"
+        file_size = 999
+        model, _ = create_random_file(file, file_size)
+        sig_path = signature_path(model)
+        result = Serializer.serialize_v0(model, 0, sig_path)
+        # NOTE: we want to also test a chunk size larger than the files size.
+        for c in range(1, file_size + 1):
+            r = Serializer.serialize_v0(model, c, sig_path)
+            assert (r == result)
+        cleanup_model(model)
+
+    # File serialization returns the same results for different file names.
+    def test_different_filename(self):
+        file = "model_file"
+        data = b"hellow world content"
+        model = create_file(file, data)
+        sig_path = signature_path(model)
+        r0 = Serializer.serialize_v0(model, 0, sig_path)
+        cleanup_model(model)
+
+        file = "model_file2"
+        model = create_file(file, data)
+        sig_path = signature_path(model)
+        r1 = Serializer.serialize_v0(model, 0, sig_path)
+        cleanup_model(model)
+
+        assert (r0 == r1)
+
+    # File serialization returns a different result for different model
+    # contents.
+    def test_altered_file(self):
+        file = "model_file"
+        file_size = 999
+        model, content = create_random_file(file, file_size)
+        sig_path = signature_path(model)
+        result = Serializer.serialize_v0(model, 0, sig_path)
+        for c in range(file_size):
+            altered_content = content[:c] + bytes([content[c] | 1]) + \
+                content[c:]
+            altered_file = file + (".%d" % c)
+            altered_model = create_file(altered_file, altered_content)
+            altered_sig_path = signature_path(altered_model)
+            altered_result = Serializer.serialize_v0(altered_model, 0,
+                                                     altered_sig_path)
+            assert (altered_result != result)
+            cleanup_model(altered_model)
+        cleanup_model(model)
+
+    # TODO(#57): directory support.